The present invention relates to a computer system comprising a computer which executes an application program.
Technology called CDP (Continuous Data Protection) is known. CDP-related technology, for example, is disclosed in Japanese Patent Laid-open No. 2005-18738. According to this Japanese Patent Laid-open No. 2005-18738, a journal is created and stored each time a storage system receives a write request from a host computer and writes data to a logical volume, and when a restore is performed, a journal is used to restore data from an arbitrary time to a logical volume.
For example, there is a computer system, which comprises a storage system comprising a plurality of logical volumes; a computer (hereinafter, job host) for executing one or more job application programs (hereinafter, job AP) using the plurality of logical volumes; and a computer (hereinafter, storage management host) for managing the plurality of logical volumes. In an environment in which this computer system is operated, there may be an administrator (hereinafter, storage administrator), who uses the storage management host to manage the plurality of logical volumes, and an administrator (hereinafter, AP administrator), who manages a job AP, which is being executed by the job host.
In this computer system, the storage management host can send to a storage system a restore indication, which specifies the storage administrator's arbitrary RPO (Recovery Point Objective) time, in accordance with an operation by the storage administrator. In this case, the storage system, in response to this restore indication, can restore data of the specified RPO time to a logical volume.
Here, for example, when a failure occurs in the storage system while a job is being performed, and the storage administrator restores data in the storage system up to an arbitrary RPO time, the AP administrator does not know what data was restored. Further, the AP administrator also does not know if the restored data affects an AP, which he is managing. When these points need to be specified, the AP administrator must query the storage administrator.
Further, for example, the RPO time does not necessarily have to be an arbitrary time of the storage administrator, but can also be an arbitrary time decided by the AP administrator. However, the AP administrator is not aware of the utilization status of a logical volume by a job AP other than the job AP that he himself is managing. For example, there are times when a logical volume being used by an AP administrator-managed AP is also being used by a job AP other than the job AP that the AP administrator is managing himself. Thus, it is difficult for the AP administrator to decide an RPO time.
The above-mentioned problems can also exist when an AP is not a job AP.